A five-year continuation is proposed to expand a prospective study relating the prevalence and incidence of metabolic and immunologic indices of islet cell damage and/or glucose intolerance to the maternal and offspring HLA phenotypes in a cohort of 600 adolescent children with rubella embryopathy at four residential schools for the deaf. The study thus exploits the naturally occurring paradigm of isletopathy in the 8000 children with congenital rubella who were born during the 1964-65 pandemic to test the general hypothesis that diabetes can arise from a genetically determined adverse host response to viral infection either in the affected individual or in the mother during pregnancy. The study will further define the rubella syndrome and test existing evidence that intrauterine maternal factors can be important in the etiology of diabetes, as well as confirm preliminary demonstrations that the incidence of diabetes, glucose intolerance, thyroid and parietal cell auto-antibodies and the HLA B15 and B40 specificities are all increased in patients with the congenital rubella syndrome. The proposed study should provide significant new information about the prevalence of diabetes-prone genotypes in the general population, racial differences in the frequency of these genes and the clinical significance of specific HLA haplotypes in the expression of diabetes in the congenital rubella syndrome. This project will involve the collaboration of scientists in the Departments of Human Genetics, Medicine, and Surgery at the Medical College of Virginia with scientists in the Department of Pediatrics at the University of Virginia and the students, faculties and administrations of four residential schools for the deaf in Maryland, Virginia, and the District of Columbia.